Live data from Hacker News

“I've had to relearn coding to get through the new interviews”

news.efinancialcareers.com

411–420 of 731 posts

Re: “I've had to relearn coding to get through the new interviews”

#411
post #373

> If I'd had the opportunity to read-up on it, I would have been fine. This highlights the paradox of Hackerrank-style tests. They would make sense in an age where the overhead to refresh your knowledge about something took time and effort: Find a relevant paper book, read it, perhaps type-out the examples. But in that age. where they would make sense, we didn't have the technology to make them available to the masse…

The point of hacker rank isn't to tell apart the code-monkey from the expert who is rusty. It's to find the people who are willing to spend hundreds of unpaid hours grinding leetcode. The thought process is that you only need two things to become a successful developer. Baseline critical thinking skills and the willingness to spend 60 hours a week reading up on best practices. If you can pass leetcodes interviews you've probably got both those. Figuring out if someone is an expert coder is hard, figuring out if they have these skills that you believe are required to become an expert coder is easier.

Re: “I've had to relearn coding to get through the new interviews”

#412

Earlier quoted context omitted.

You're right, and as a result I try to work no more than 4 hours a day, preferably less. I take every possible opportunity to extend timelines, create buffer, and reduce scope. I'm highly incentivized to farm as much time as possible, so that's what I do. I try to target obscure projects that I already know how to do, where there's an assumption that nobody really knows how to do it because it's obscure. Complexity a…

This sums up perfectly why I’m pro universal basic income. How many people do this? Think of all the collective brain cycles wasted. We’re stealing from our future. You’re clearly smart and capable. Work you don’t find rewarding stole your motivation. If you could work on projects that reflect the world you want to live in you would create something much more valuable top society than whatever it is you’re half assin…

I agree, but I think the reason we're unlikely to see something like this no matter how much evidence we present that it is better for everyone is that people with money and power generally got there because they craved money and power. Owning people's time, even if that time isn't being spent productively for them, is as close as they can get to owning people themselves and they desperately crave the power to own people.

Re: “I've had to relearn coding to get through the new interviews”

#413
post #370

What if we form a group/site that refuses to do leetcode interviews? This seems like a classic coordination problem. If enough people refuse, companies will have their talent pool limited. We could use this group to help match companies and developers who don't want to leetcode. If it's true that leetcode filters out great developers, than companies who 'can't find talent' should be able to use our group to find some…

You are describing a labor union.

I agree, I avoided using the word union.

In an ideal world, this would be a union.

In (what I worry might be) the real world, the word union conjures unhelpful barriers to action.

I'm less interested in fighting idealogical battles. I'm more interested in achieving a practical goal - getting a large group of people to say no to leetcode.

My guess is more people will say no to leetcode if they think of it as a general group anyone can join. This allows a larger amount of people to ignore idealogical barriers they might have to joining. So for now it's a "group/job board".

Re: “I've had to relearn coding to get through the new interviews”

#414

I've just landed a job at a FAANG company. Some random subjective comments based on my experience: - it was hard work - there was no way I could have passed these interviews without all the work. - my past experience and education barely helped me for the interview - it's not enough to be able to solve the problems, you need to be able to solve them on the spot, without much thinking, in an interview context - I pass…

One big problem is that companies (especially FAANG) are wasting time having very senior devs do Leetcode-style shit, which is a false negative for people that don't balance binary trees and manually regex strings all day. This especially penalizes more experienced / older devs who don't have time to LC grind (of which I'm also one), and introduces unconscious selection bias.

Certainly there are other ways to verify someone can code: code samples, talking through high level language concepts, reviewing existing code, etc.

What people should be focusing on is this - 1) Do you understand enough systems architecture / code techniques to succeed at this particular company? This can be done via mostly talking. 2) Do you have experience that will help us with our company-specific problems? Give them an actual problem you regularly deal with, and get them to attack it.

Re: “I've had to relearn coding to get through the new interviews”

#415

Earlier quoted context omitted.

> - it's not enough to be able to solve the problems, you need to be able to solve them on the spot, without much thinking, in an interview context Sounds like they're looking for assembly line programmers - crank out the code, pass it off, then on to the next piece of code without pause. This would be a nightmare for me. The high salaries and prestige of these companies definitely would not be worth it. I'm not neur…

I think what is required for the interviews is very different than the actual work you do once you are hired. It 's also very team dependent. The reason why you're expected to be fast during the interviews is that you're competing with people who are fast (either because they are smart, or because they've practiced a lot). That being said, this type of interviews is clearly not for everyone. There are tons of skilled…

> I think what is required for the interviews is very different than the actual work you do once you are hired

I'm sure that's true once you achieve seniority. I'd wager that the first couple of years, at least, you're not solving problems, but implementing solutions designed by higher ups. You need to do this fast to be useful.

Re: “I've had to relearn coding to get through the new interviews”

#416
post #372

Earlier quoted context omitted.

>until I make it into such an "endgame" company. Unless you're close to retirement and don't really need the money, it's unlikely you'll know when you're there. Companies change and/or get into trouble, managers change, projects get canceled, etc. I've been at companies I spent quite a few years at and I still came close to quitting for various reasons (and eventually did). I'd submit there is no predictable endgame…

You're right that "endgame" is different from person to person, and also in the same person's lifetime. My current company was something I would have considered endgame at the start of my career. Now it's not. A company I'd consider to be currently endgame (say, FAANG), may not be 10 years later. For the sake of discussion, as I mentioned in my original comment, I would simply define an endgame company as one where I…

I confess that, as someone who has never been primarily a developer (done a number of different things over time), I find this whole leetcode thing beyond strange. I have never in my life studied for interviews other than reading up on a specific company I'm interviewing with. ADDED: And every job since the one out of grad school decades ago has been directly through someone I knew.

Re: “I've had to relearn coding to get through the new interviews”

#417
I wonder how many HNers are old enough to remember when Joel Spolsky shook up a lot of thinking about software hiring at the time (~2006). He had two major insights, one of which has become lost to time unfortunately and the other has mutated into an abomination over time:

- Hire based on "smart and gets things done"

- FizzBuzz

Prior to Spolsky software interviews were much more like a traditional job interview: resume, CV, talk to a manager, a bunch of questions about experience etc. Things like N years experience in language X where hugely important. The idea that you should hirer people that were "smart" and "got things done" was mind blowing at the time.

Unfortunately FAANG seems largely focused on hiring for their definition of "smart" and has virtually no interest in if you can get things done. I've learned when interviewing at these places to never mention any past experience I've had solving a problem since it always seems visibly hurt my performance.

And VC funded startups have really doubled down on the "not smart" and "get things done" category. All of the sub 1k employee startups I've worked at or talked to in recent years seem to penalize thinking about or understanding the problems your working on, and mostly care that you're constantly scrambling to get things done, even if those things are useless.

So sadly, "Smart, and get things done" has been completely lost to time. FizzBuzz on the other hand...

The entire point of FizzBuzz was to make sure that people could code. It was supposed to be very easy, and completed in a very short period of time. Again, the problem in 2006 was that there were many corporate programmers who only knew how to basically refactor legacy Java code. If you had to start from zero, which most startups did back then, you needed someone that could build a program from nothing and think about problems computationally. FizzBuzz was meant as a sanity check, not a stress test.

This has morphed into a bizarre paranoia that "most coders can't code!" Now the point of FizzBuzz wasn't to replicate the exact problem, but to come up with some other quick, easy problem just to verify that the candidate does, in fact, know how to code. But hiring managers had no imagination so they just replicated the exact example Spolsky used. This lead to candidates memorizing the problem as part of interview prep. This of course lead to companies developing harder problems, which lead to candidates studying more. When this combined with Google's famous algorithm questions (originally to check if, you know, you did in fact understand computer science) we ended up with the leetcode nightmare we have today.

It's tragic since many of the "smart and gets things done" programmers today struggle to pass interviews. But perhaps this is because companies don't really want that anymore.

Re: “I've had to relearn coding to get through the new interviews”

#418
post #406

Earlier quoted context omitted.

It's because the business people got interested and are increasingly taking leadership roles. As an engineer who has shifted over to the product management side I can safely say business people suck fun out of stuff for a living. Margins, cost savings, metrics to manage, each and every slow step of creating a new, less fun, normal erodes a culture built on making "Something cool" or "Just for fun". Worse yet is when…

Agile/Scrum was the biggest bait-and-switch con out there - developers thinking the methodologies would empower them while in fact being a vector for the worst excesses of micromanagement. I feel less empowered now than I did in my earliest days as a junior dev.

Yep, I call it collectivised micromanagement in an attempt to commoditize development.

Re: “I've had to relearn coding to get through the new interviews”

#419
post #416

Earlier quoted context omitted.

You're right that "endgame" is different from person to person, and also in the same person's lifetime. My current company was something I would have considered endgame at the start of my career. Now it's not. A company I'd consider to be currently endgame (say, FAANG), may not be 10 years later. For the sake of discussion, as I mentioned in my original comment, I would simply define an endgame company as one where I…

I confess that, as someone who has never been primarily a developer (done a number of different things over time), I find this whole leetcode thing beyond strange. I have never in my life studied for interviews other than reading up on a specific company I'm interviewing with. ADDED: And every job since the one out of grad school decades ago has been directly through someone I knew.

You're not the only one. My wife finds the whole thing bizarre. As do my non-techie friends.

My wife's best friend thought it stupid (actually, I agree - it is stupid). Then she started dating a SWE, who is apparently also leetcoding in his spare time.

Re: “I've had to relearn coding to get through the new interviews”

#420

I've just landed a job at a FAANG company. Some random subjective comments based on my experience: - it was hard work - there was no way I could have passed these interviews without all the work. - my past experience and education barely helped me for the interview - it's not enough to be able to solve the problems, you need to be able to solve them on the spot, without much thinking, in an interview context - I pass…

> - it's not enough to be able to solve the problems, you need to be able to solve them on the spot, without much thinking, in an interview context Sounds like they're looking for assembly line programmers - crank out the code, pass it off, then on to the next piece of code without pause. This would be a nightmare for me. The high salaries and prestige of these companies definitely would not be worth it. I'm not neur…

One of my big impressions of Go is that it's designed to facilitate exactly that. I've also heard comments in that vein.
Post reply on HN